Skip to main content

Full text of "Thermal Analysis and Testing of Fastrac Gas Generator Design"

See other formats





AIAA 98-3366 

Thermal Analysis and Testing of 
Fastrac Gas Generator Design 

H. Nguyen 

NASA Marshall Space Flight Center, 

Huntsville, AL 35812 




34th AIAA/ASME/SAE/ASEE 

Joint Propulsion Conference & Exhibit 

July 13-15, 1998/Cleveland, OH 



For permission to copy or republish, contact the American Institute of Aeronautics and Astronautics 
1801 Alexander Bell Drive, Suite 500, Reston, VA 20191 



Thermal Analysis and Testing of 
Fastrac Gas Generator Design 

H. Nguyen* 
NASA Marshall Space Flight Center, Himtsville. x\L 35812 

Abstract 

The Fastrac Engine is being developed by the Mcir- 
shall Space Flight Center (MSFC) to help meet the 
goal of substantially reducing the cost of access to 
space. This engine relies on a simple gas-generator cy- 
cle, which burns a small amount of RP-1 cind oxygen 
to provide gas to drive the turbine and then exhausts 
the spent fuel. 

The Fastraic program envisions a combination of 
analysis, design and iot-fire evaluation testing. This 
paper provides the supporting thermal smalysis of the 
gas generator design. In order to ensure that the de- 
sign objectives were met, the evaluation tests have 
stcu-ted on a component level cind a total of 15 tests of 
different durations were completed to date at MSFC. 
The correlated thermal model results will also be com- 
pared against hot-fire thermocouple data gathered. 

Introduction 

DURING the past several years, increasing empha- 
sis has been given to the development of a low- 
cost space transportation system.' Two key areas with 
potential for limiting the cost of future space tranpor- 
tation sustems are efficient engine development cind 
optimal utilization of inexpensive propellants, such 
as the LOX(liquid oxygen) /RP-1. To make further 
progress In the above key areas, in 1996 MSFC spon- 
sored the Low-Cost Boost Technology Project,^ the 
centerpiece of which is the development of a econom- 
ical reusable engine derived from previous technology 
programs for turbopump"'"^ cuid chamber^to serve as 
the main propulsion system (MPS) for the X-34 vehi- 
cle. 

The X-34 MPS features, among many others, a con- 
ventional gas-generator cycle and simple robust design 
using commercial off-the-shelf components to encour- 
age nontraditional vendors and smcill corporations to 
introduce commercial design and manufacturing pro- 
cesses advantageous for space transportation. Another 
goal set for the MPS is its minimal maintenance to 
meet operability requirements. Interested readers may 
consult other documents^ "'^ for more details about 
the design and development of the MPS. The rest of 




-1 Chamber Body 

60k Fastrac 
Gas Generator 



'Senior Member AF.A.^ 

Copyright © 1998 by the American Initilute of AeronKutict and Aitro- 
nniiti<ri. Inc. No copyright it Jts»«rtcd in the United Stales under Title 17. U.S. 
Code The U.S. Government h»e iv royalty-free license to exercise all rights un- 
der the copyright claimed herein for Governmental Purposes. All other rights 
are reserved by the copyright owner. 



Fig. 1 Combustion Chamber 

this paper focuses on the Fast rac gas generator design, 
thermal analysis and testing results. 

Gas Generator Design 

The Fastrac gas generator baseline configuration 
consists of a faceplate brazed to an injector assembly 
and a combustion chamber. 

The uncooled combustion chamber is a cylinder 
whose length and diameter are 8.875 in. and 3.535 
in., respecti%ely. Turbulent mixing near the chamber 
wall is further promoted by the slotted turbulence ring 
(Figure 1), whose exact distance from the faceplate 
is going to be determined by hot-fire testing. This 
chamber is made of Hastelloy-X, a nickel-base alloy. 
The higher thermal conductivity (compared to steel's) 
of nickel provides additional design margin since local 
hot spots are better diffused. Other materials selection 
criteria for this project are 

• Low-cost and easily obtainable; 



Easily weldable; 



I OF 8 



American Institute of Aeronautics and Astronautics Paper 98-3366 




Fig. 2 Injector/Faceplate 

• Brazeable by demonstrated processes; 

• Compatible with LOX, RP-1, and the resulting 
combustion products. 

The brcized injector assembly (Fig. 2) is a single- 
piece, 304L-stainless steel manifold body brazed to a 
3.535-inch diameter faceplate constructed of oxygen- 
free, high-conductivity copper. The selected mate- 
rials met the listed criteria. Each injector element 
is composed of one pair of self-impinging RP-1 ori- 
fices shielding a single oxidizer orifice in a conventional 
fuel-oxidizer-fuel (F-O-F) triplet arrangement. Table 1 
gives the orifice diameters and other pertinent design 
parameters. 

For testing purpose, a turbine simulator (Fig. 3) 
was included to supply the back pressure. A cham- 
ber spacer (Fig. 4) was also utilized to allow tests on 
a longer chamber. An instrumentation rake at the 
exit plane contains six thermocouple probes for turbine 
inlet temperature distribution measurements in the ra- 
dial direction, and two pressure probes (Fig. 5). The 
disassembled and assembled test hardware are shown 
by Fig. 6 and Fig. 7. 

In the following sections, the objectives of thermal 
analysis, the governing equation, and solution proce- 
dure will be taken up. 

Objectives 

The thermal analysis has several primary aims: 

• Minimize the RP-1 freezing in the injector man- 
ifolding (avoid unpredictable combustion charac- 
teristics) 

• Provide adequate cooling of the injector faceplate 

• Maintain hardware structural integrity i.e. no 
melting 




Fig. 3 Turbine Simulator 




Fig. 4 Chamber Spacer 



2 OF 8 



American Institute of Aeronautics and .-Vstronautics Paper 98-3366 



Nominal Chamber Pressure, psia 


575 


Oxidizer Flowrate, Ibm/sec 


1.64 


Fuel Flowrate, Ihm/sec 


5.46 


Mixture Ratio 


0.3 


Gas Exit Temperature, °R 


1600 


Exit Temperature Profile, °R 


±50 


Number of Oxidizer Orifices/Dia., in. 


42/0.034 


Number of RP-1 Orifices/Dia., in. 


84/0.047 


LOX Mean Injection Velocity, ft/s 


90 


LOX Injection Pressure, psia 


702 


RP-1 Mean Injection Velocity, ft/s 


107 


RP-1 Injection Pressure, psia 


702 



Table 1 Design Parameters 




Fig. 5 Instrumentation Ring 

• Supply temperature distribution for the subse- 
quent stress analysis 

The stress analysis is the subject of a separate study 
and will not be reported here. 

Governing Equation 

The governing diff'erential equation* for the conduc- 
tion of heat in solid is 



pcp^ = V.(fcVT) + q. 



where 



Cp = specific heat 

k = thermal conductivity 

T = temperature 

t — time 

p = density 

Qv = volumetric rate of internal heat generation 

The specific heat Cp is a function of T, and is related 
to internal energy, U, through 




dU 

IT 



Fig. 6 Gas Generator Components 

For most practical designs using many types of 
materials and operating over a wide range of T, 
temperature-dependent and spatial variations of ther- 
mal conductivity k must be considered. The resulting 
problem is nonlinear with boundary conditions speci- 
fied on complex boundary. 

Solution Procedure 

The PATRAN^ commercial software was utilized to 
automate the tasks of modeling complicated geometry, 
imposing boundary conditions and material proper- 
ties, and post-processing a massive amount of analysis 
results for a multi-dimensional configuration. 

Figure 8 shows the solution procedure used in the 
thermal analysis. For basic mesh generation, the 
only required information is the desired element size 
and the boundary geometry for each material. PA- 
TRAN produces the nodal points, the conductors link- 
ing those nodes, and outputs the required lines for a 
SINDA thermal model. 

SINDA^" solves the nonlinear heat conduction equa- 
tion shown earlier using a lumped pareimeter finite- 
difference method where the geometry to be modeled 
is divided into lumps of mass called nodes that are 
connected to each other with PATRAN-generated con- 
ductors. The output file containing the temperature 
distribution from the SINDA calculation can be post- 
processed by PATRAN after each successful run. 

Figure 9 shows a two-dimensional axisymmetric grid 
of the gas generator generated by PATRAN. 

Hot-fire Testing 

The primary goal of this component-level test is to 
evaluate and improve, if necessary, the baseline gas 
generator design (Fig. 7) for the Fastrac engine. More 
specifically, the tests are to describe 



3 OF 8 



American Institute of Aeronautics and Astronautics Paper 98-3366 




Fig. 7 Test Article 




c;;;^" Geometry 



celerometers, propellant inlet pressures, fuel injector 
manifold temperature, and external skin temperature. 
No thrust measurement was attempted. Test data 
were made available to analysts as soon as the test 
ended. 

Testing'^ was performed at MSFC Test Stand 116. 
.\ total of 15 tests were completed to date. Tables 2- 
5 show the duration, chamber pressure, and mixture 
ratio measured for each test. 



Fig. 8 Thermal Analysis Procedure 

• Hot-gas temperature distribution at the inlet of 
the turbine simulator 

• Performance based on measured flow rates of pro- 
pellants, chamber pressure 

The target for each of the above is outlined in Ta- 
ble 1. 

The propellant flow was regulated using cavitating 
Venturis. These devices provide a constant mass flow 
for a given inlet pressure and density. Propellants are 
provided in a pressure-fed mode. 

Instrumentation on the test stand monitored all crit- 
ical operating parameters. Test stand instrumentation 
included propellant tank pressures, feedline pressures 
and temperatures, and propellant flow rates. En- 
gine instrumentation included chamber pressure, ac- 



Test Number 


01 


02 


I 03 


04 


Test Duration, sec. 


7.5 


7.5 


5.8 


13.3 


Mainst. Dur., sec. 


0. 


0. 


2.0 


10.0 


Chamber Press., psig 


85 


1.35 


475 


470 


Mainst. Mix. Ratio 


- 


- 


0.29 


0.29 



Table 2 Tests 1-4 Summary 



Test Number 


05 


l06 


07 


08 


Test Dur.. sec. 


3.3 


63.8 


153.8 


103.8 


Mainst. Dur.. sec. 


1.0 


6.0 


150.0 


100.0 


Cham. Pr., psig 


477 


534 


530 


535 


Mainst. M.R. 


0.30 


0.30 


0.30 


0.31 



Table 3 Tests 5-8 Summary 

The maximum hot-gas temperature gradient mea- 
sured at the turbine simulator inlet was 64 deg. F. 
The minimum was 28 deg. F. The ±50 deg. F target 
(Table 1) was reached for most tests. 



1 OF 8 



American Institute of .Aeronautics and .Astronautics Paper 98-3.366 



Test Number 


09 


10 


11 


12 


Test Dur., sec. 


153.9 


7.6 


63.8 


23.8 


Mainst. Dur., sec. 


150.0 


2.0 


60.0 


20.0 


Cham. Pr., psig 


520 


475 


483 


480 


Mainst. M.R. 


0.30 


0.24 


0.25 


0.25 



Table 4 Tests 9-12 Summary 



Test Number 


13 


14 


15 


Test Duration, sec. 


63.9 


63.8 


33.9 


Mzunstage Dur., sec. 


60.0 


60.0 


30.0 


Chamber Press., psig 


470 


454 


475 


Mainstage Mix. Ratio 


0.25 


0.33 


0.25 



Table 5 Tests 13-15 Summary 




Fig. 9 2-D Patran Grid 

Calculation Results and Discussion 

In order to predict the solid RP-1 thicicness in the 
fuel manifold (Fig. 2), a simple one-dimensional ther- 
mal model'" of the annulus wall separating LOX and 
RP-I passages was built (Fig. 10). Results indicated 
that a solid film, averaging 0.009 inch, could develop. 
Flow area is reduced up to 20 percent as a result. .An- 
other design chcirt estimated a solid RP-1 layer 0.006 
to 0.008 inch thick for the present operating condi- 
tions. Concerns about RP-1 freezing have led MSFC to 
instrument additional thermocouples to monitor RP-1 
bulk temperature. Test data showed, as expected, that 
bulk RP-1 temperature inside a fuel annulus remained 
above 80 deg. F, exceeding the -50 deg. F freezing 
mark (Fig. 11). 

For the pre-test calculations shown below, the hot- 
gas environment''' was numerically simulated by a 
separate in-house study. The resulting set of convec- 
tive heat transfer coefficients were input as boundary 
condition for the SINDA model. 

The injector faceplate (Fig. 2) is cooled during en- 
gine run by employing the two RP-1 jets to shield 
against the single LOX jet. .A. ma.ximum surface tem- 
perature of about 200 deg. F was calculated. No 
data is available for model correlation but the post- 
test inspections revealed no damage. Extra margin 
was provided by the insulation effect of soot layer de- 



RP-I 
552 R 




LOX 
233 R 



304L Plate 
Frozen RP- 1 



Fig. 10 1-D Model 



positing during test. 

The two-dimensional model (Fig. 9) was exercised 
to give temperature distribution on the combustion 
chamber wall. This is an area of concern because it is 
uncooled zind a proper design for a heat shield brzuJcet 
mounted on the externid surface has to account for 
this hot boundary. 

Figure 12 shows a sketch of the test article with 
externd thermocouples. Shown by Figure 13 and Fig- 
ure 14 are the test data and predictions. At its worst, 
the correlated model's results were within 50 degrees of 
test data. The maLximum temperature, approximately 
1100 F, was about the same for chamber locations that 
aie 2.5 inches apart. This is probably due to the tur- 
bulent mixing process of the combustion products with 
the propellants having been completed, giving rise to 
flow uniformity the rest of the way. 

Figure 15 shows the ajialysis results compared to the 
data at two locations near the faceplate. The max- 
imum temperatures are different because unlike the 
situation in the preceding paragraph, mixing was in- 
complete. 

Conclusions 

The baseline design of the Fastrac gas generator was 
analyzed amd the results were compared against appli- 
cable hot-fire data. 

The maximum temperature on the chamber externeJ 
surface was about 1100 deg. F. There were reason- 
able agreements between data and correlated numeri- 
cal model. 

Annulus flow area could be reduced by up to 20 
percent although the RP-1 bulk temperature remained 
well above freezing during testing period. 

The faceplate triplet design was adequately cooled, 
having suffered no erosion in amy test thus far. 

Acknowledgements 

This work was supported by the N.\SA Office of 
.\eronautics and Space Transportation Technolog>' 
(OASTT), Washington, D.C. 



5 OF 8 



A.MERicAN Institute of Aeronautics and .Astronautics Paper 98-3366 



120.0 




40.0 
Time (seconds) 

Fig. 11 RP-1 Manifold Temperature 



80.0 



8 T/Ci m swn* ptan*. .8' from ftang* in tha lolowing cktck potitians (kMking ai ini«clor): 
Th«s« m«wurwn«nts ar* tor mapping out any infactor hot spots in 
local iorti similar to what haa baan saan in Iha past. 



T733a 


1:00 


T7331 


2:00 


T7339 


4:00 


T732I 


SOO 


T7336 


8:00 


T7324 


10:00 


T7337 


It 00 


T73I2 


12 00 


T3001 




jccet iuftice iP»r" 


•m- 


— p fiU 


■ti 




Fig. 12 Test article on test stand (Black dots denote thermocouple locations) 



References 

'Anon., "'N.\SA Studies Access to Space." Advanced Tech- 
nology Teann Volumes 1-4, 1993. 

^Anon., •'Low-Cost Boost Technology Project," World Wide 
Web page address http://tvww.ies.msfc.nasa.gov/lcbt, 

■'Garcia, R., ''Computational Fluid Dynamics .Analysis in 
Support of the Simplex Turbopump Design,"' iVASA CP-3302, 
May 1994. 

■"Garcia, R., "Fluid Analysis of Pump Manifolds Designed 
for Cost," JANNAF Propulsion Meeting, 1996. 

'Sparks, D., "Ablative Combustion Chamber Liner Feasi- 
bility Study," NASA TM-i08470, 1994. 

^Sgarlata, P. and Winters, B., ■'X-34 Propulsion System De- 
sign," AIAA Paper 97-3304 , 1997. 

^Anon., "'Low-Cost Boost Technology Design Binder," Tech. 
Rep. Vol. 1, Book 3, NASA, 1997. 

*Eckert, E., ".Analysis of Heat and Mass Transfer," 1987. 



^Anon., "P3/P.ATRAN User Manual," Tech. Rep. Vol. I, 
1993. 

'"Behee, R., "SINDA/G Tutorial Guide," Tech. rep., 1996. 
"Sanders, T., "Private Communication," 1997. 
'^Luong, v., "Private Communication," 1997. 
'■'Canabal, F.. "Private Communication," 1997. 



6 OF 8 



American Institute of Aeronautics and Astronautics Paper 98-3366 



Chajnber Surface T7323 



1500.0 



— 1000.0 i- 



500.0 




0.0 



■0 Test 13 
o Node 3987 

• Node 3988 

* Test 14 
-i Test 15 

* Test 7 
> Node 9470 

• Node 9471 
- Test 9 



50.0 



100.0 
Time (seconds) 



150.0 



200.0 



Fig. 13 Comparison vs. T7323 thermocouple data 



1500.0 



Chamber Surface T7322 







500.0 



50.0 



100.0 
Time (seconds) 



-» Test 13 
= Node 3961 

• Node 3962 
« Test 14 
■•Testis 

• Test 7 

» Node 9453 

• Test 9 



150.0 



200.0 



Fig. 14 Comparison vs. T7322 thermocouple data 



OF 8 



American Institute of Aeronautics and Astronautics Paper 98-3366 



200.0 



150.0 



- 100.0 



50.0 



-50.0 



-100.0 



0.0 



60K Fastrac Gas Generator 

Faceplate Thermocouple T7312 



3 3 a a 




-oT7312(T«sl30) 

a Prediction 
-»T7312(Teal2g) 



50.0 



100.0 
Time (seconds) 



150.0 



200.0 



1500.0 



— 1000.0 



500.0 



0.0 



0.0 



60K Fastrac Gas Generator 

Faceplate Themiocouple T7321 




oTTM! (TMt30) 
-<iT7321 |Tm122) 
oPradlcPon 



50.0 



100.0 
Time (seconds) 



150.0 



200.0 



Fig. 15 Comparison vs. T7312 and T7321 thermocouple data 



8 OF 8 



American Institute of Aeronautics and Astronautics Paper 98-3366